1. Field of the Invention
The present invention relates to image processing apparatuses and, particularly, to an image processing apparatus that recognizes the state of an image to perform a correction process including upside/downside identification and skew correction.
2. Description of the Related Art
Conventional analog copiers have been changed to digital copiers using digitized image data to produce copy. Such digital copiers have an increased affinity with other apparatuses handling digital image data. With not only the functions as copiers but also various functions, such as facsimile, printer, and scanner functions, being combined, these copiers have been not merely taken as simple digital copiers but have been called digital multifunctional peripherals (MFPs). With an increased capacity and a decreased cost of memory, a function of changing a paper document to digital image data has advanced. Technologies associated with MFP devices have also been advancing. High-speed network has been widespread, the process performance of Central Processing Units (CPUs) has been increased, and compression technologies associated with digital image data and others have been advancing. Accordingly, the variety of functions incorporated in the MFP devices has also been increased. Furthermore, in recent years, the variety of use of the MFP devices has been increased. In an increasingly-used scheme, digital image data is accumulated and stored inside the MFP device and the data is output again when this information is required.
MFP devices have been diversified into a small size to a large size. There is a small MFP device set on a personal computer (PC) side and allowing an operator to easily use facsimile, printer, and scanner functions. Also, there is a middle-sized MFP device shared among a plurality of persons for each department and section and allowing sorting, punching, and stapling functions to be used with productivity to some extent. In a section centrally in charge of copy-related tasks within a company or a company whose core business is copy-related tasks themselves, a large MFP device with high productivity, high quality, and multifunction is used.
Although there are functions shared among every class, there are also functions strongly required for each class. For example, large-sized MFP devices require post-processing on paper after plotting, such as punching, stapling, and paper folding, and electronic filing together with copying tasks. Small-sized MFP devices require expansion of Internet fax, PC fax, and the like, and high-quality image printing on dedicated paper for personal use. For each of such diversified MFP devices, a system with a set of functions required for each class is constructed.
Meanwhile, in a reading device of an MFP device, a document may be read in a tilted state. For example, the document may be tilted when the document is conveyed by using a document feeder (DF) often used at the time of printing a large amount of document. As a result, electronic data of the read image is tilted, causing an image printed on paper to be also tilted. When digital data is subjected to optical character recognition to be converted to text data, such a tilt degrades a recognition rate. Moreover, when a user sets the document without being aware of which side is upside or downside of the document, the upside and downside of electronic data is as being read, which causes a mixture of upside-down printing results. When the MFP devices are used for business purpose, to correctly arrange the upside and downside of the document, the user has to arrange the upside and downside by himself or herself.
To solve such inconveniences, there is an MFP device provided with a function of recognizing the upside and downside of a document for correction and detecting a tilt of a document for correction to eliminate the user's need of being aware of a tilt of the document. Also, the upside and downside is automatically detected and rotated by a PC optionally connected to the MFP device, and a tilt angle of the document is determined to correct the tilt. In the following, examples of conventional technologies associated with such MFP devices are explained.
Japanese Patent Application Laid-Open No. 9-200507 discloses an “image reading apparatus” in which a tilt is detected in a document read by an automatic document feeder (ADF) for correction. Tilt correction is performed by an ADF mechanism on a tilt of the document with respect to a scanner unit. The document is sequentially converted to image information for each line in a main scanning direction, and an automatic document feeder mechanism (ADF mechanism) is used to move the document in a sub-scanning direction. By using a carrier sheet having an identification frame for detecting a tilt angle θ of the document with respect to the scanner unit, the tilt angle θ in the document scanning direction is detected and then stored. By using the stored tilt angle θ, the position of the image information is corrected. With the tilt angle θ in the document scanning direction due to a mechanical shift of the scanner unit or the like being detected and retained, the tilt angle θ of the image data is subjected to tilt correction for output when the document is again read.
Japanese Patent Application Laid-Open No. 10-327310 discloses an image reading apparatus in which image data corrected according to a tilt of the document can be output. This image reading apparatus can prevent a loss of image data near the edge of the document in image data to be output, thereby increasing convenience of image data to be output. Image data including the entire image surface of the document is read and stored. A rectangle that specifies a mounting position of image data and a document of a standard size is displayed. The image data is then transferred from an image reading and storage unit to an image displaying unit. The image data is then corrected according to a tilt of the edge portion of the document crossing one side corresponding to one side of the rectangle. The image data corrected by a data correcting unit is then output.
Japanese Patent Application Laid-Open No. 11-0717892 discloses an “image processing apparatus” in which a correction process is performed according to a tilt of a document surface with respect to a reading unit, thereby obtaining an excellent image without skewness. A read surface of a document and the reading unit that reads the read surface are tilted by a specified angle, and image information entered from an input unit is processed. A correction at the time of image processing is performed according to a specified angle, which is a tilt between a document movement reading direction and the reading unit. A correction process according to a tilt between the document surface and the reading device is performed.
Japanese Patent Application Laid-Open No. 2001-358914 discloses an “image reading apparatus” in which skew detection is included in a reading device itself, and a correction process is performed based on the detection results. A skew of a document is determined based on an output of a line sensor that reads an image on a document, thereby correcting a skew of image data. In the image reading device, when a charge-coupled device (CCD) receives reflected light of light emitted toward the document and reads the image on the document, an image processing unit performs a document-size detection process in which a position and width of the document is detected from an output of the CCD. Based on the detection results, a skew, which is a tilt of the document, is determined by a CPU. Based on the determined skew, the tilt of image data of the read document is corrected by the image processing unit under the control of the CPU. Therefore, without requiring provision of a new detecting unit, a skew of the document can be detected based on the reading results of the CCD that reads the image on the document, and the tilt of the image data can be corrected according to the skew of the document.
Japanese Patent Application Laid-Open No. 2002-111988 discloses an “image processing apparatus” in which a second image processing unit that performs image processing on accumulated and stored image data is provided. This can easily achieve high-quality image processing, and increases image processing speed. Also, this can achieve high-quality image processing and increased processing speed over the entire processing system. A first image processing unit is provided that processes image data provided from an image input unit, an FCU, and a PC. An image bus management CDIC that collectively manages image data transmission and reception with respect to IPP over a data bus is provided. Also provided is a memory management IMAC that collectively manages transmitting and receiving accesses with respect to memory over the data bus. Furthermore, provided is the second image processing unit in which an access to the memory is managed by the memory management IMAC and image data processing can be performed by accessing data on the memory. This image processing apparatus also includes a dedicated path for a second IPP2 to directly access the memory for reading and writing data.
Japanese Patent Application Laid-Open No. 2003-281469 discloses a “document image processing method” in which a tilt of an image and a document direction can be determined irrespective of a brightness-reversed state in which a bright character is present in a dark background on an image with multilevel image data, and effective information as information for character recognition processing is output. A gray image is generated from an input multilevel image to generate a binary image A. Then, a tilt of the image and a direction of the image are detected to determine whether the tile and direction are successfully detected. When detection is successful, the angle of the image is corrected. When detection is failed, on the other hand, a gray image with brightness being reversed is generated to generate a binary image B. Then, a tilt and direction are detected again to generate an angle-corrected image. With this, the tile and direction of the image can be accurately detected and corrected irrespectively of the presence or absence of brightness reversion.
However, conventional image correcting methods and others have the following problems. When digital image data accumulated and stored inside an MFP device is again output, it is difficult to use the digital image data with optimum image quality according to the use purpose. For example, digital image data stored at the time of copying is transmitted through facsimile, if skew correction is performed to enhance a character portion, the image quality may be inappropriate for facsimile. Also, if a high-precision correcting unit is provided to an MFP device or an optional PC, an image correcting process optimum to all use purposes can be performed, but this merely costs users who do not aim at such use purposes and becomes a burden to the users. Moreover, for correction process other than skew correction, it is impossible for the conventional methods to achieve an optimum image according to the use purpose.